Transforming growth factor-fl1 inhibits casein secretion from ...

Biochem. J.

333 3

innGetBian Great Britain) (1994) 304, 333-336 3-3 (Printed (rne

Bice.J 19)34

RESEARCH COMMUNICATION

Transforming growth factor-fl1 inhibits casein secretion from differentiating mammary-gland explants but not from lactating mammary cells Allan W. SUDLOW,* Colin J. WILDEt and Robert D. BURGOYNE*t *The Physiological Laboratory, University of Liverpool, P.O. Box 147, Liverpool L69 3BX, and tHannah Research Institute, Ayr KA6 5HL, Scotland, U.K.

Transforming growth factor-,f (TGF/3) is important in the maturation and function of the mammary gland and is present in milk. We have examined whether, in addition to inhibiting lactogenesis, TGF/3 exerts acute regulatory effects on lactating mammary cells. The isoform TGFfl1 at 5 and 50 ng/ml suppressed the onset of lactation and the subsequent production of fi-casein by differentiating mouse mammary explants from preg-

nant mice. By contrast, it did not inhibit protein synthesis

INTRODUCTION

effect of TGF/? on casein secretion from mammary-gland explants (Robinson et al., 1993) did not directly address the issue of acute regulation of casein secretion. Recently, it has been shown that TGF/31 is present in milk from lactating mice (Letterio et al., 1994), raising the possibility that TGF/3 could act as an autocrine regulator of casein synthesis or secretion. Inhibitory feedback regulation of milk protein synthesis and secretion from lactating mammary cells (Blatchford and Peaker, 1982) is thought to involve a milk protein termed 'FIL' (feedback inhibitor of lactation) (Wilde et al., 1987, 1994). Like TGF,/, FIL can exert chronic autocrine effects on mammary-gland differentiation (Wilde et al., 1990). FIL has also been shown to inhibit acutely constitutive casein synthesis and secretion by lactating mouse mammary epithelial acini (Renninson et al., 1993). The data presented here suggest that, while developmental effects ofTGF,81 can modify casein secretion, unlike FIL, TGFfl1 is not an acute regulator of casein production by lactating mammary cells.

In the past few years, results from several studies have implicated the family of transforming growth factor-fl (TGF,) isoforms in the maturation and function of the mammary gland. Levels of mRNAs for the three mammalian isoforms of TGF/3 (denoted ,B1, /12 and fl3) peak during pregnancy, decrease during lactation and become elevated again during mammary-gland involution (Robinson et al., 1993; Streuli et al., 1993). Introduction of implants containing the three TGF,l isoforms into the mammary glands of subadult virgin mice has been shown to inhibit ductal development (Robinson et al., 1991). When TGF,/1 was overexpressed in the mammary gland of transgenic mice, pregnancyinduced lobulalveolar development and milk protein production were impaired (Jhappan et al., 1993). In addition to the inhibitory effects of TGF,8 on mammarygland growth and development, results from in vitro experiments have demonstrated a variety of effects of TGF, on mammary epithelial-cell function. For example, TGF/31 has been shown to induce expression of a number of extracellular-matrix (ECM) proteins by human mammary epithelial cells in culture (Stampfer et al., 1993). This ability of TGF,81 to stimulate ECM formation may be under feedback regulation, as TGF,/1 expression in mouse mammary epithelial cells was strongly reduced by the presence of ECM components (Streuli et al., 1993). It has been also shown that the three isoforms of TGF,3 inhibit ,8-casein production in a mammary epithelial cell line (HC 1) and in mammary-tissue explants from mid-pregnant mice (Mieth et al., 1990; Robinson et al., 1993). Robinson et al. (1993) reported that TGF,8 did not appear to affect the morphological development of mammary explants from mid-pregnant mice, but inhibited casein synthesis and release at the post-transcriptional level. It is not known whether TGF,8 exerts acute regulatory effects on differentiated lactating mammary cells. Earlier work on the

or

secretion from acini isolated from lactating-mouse mammary gland or protein secretion from explants from lactating mice. These data indicate that TGF/ inhibits the onset of casein secretion, but is not an acute regulator of casein synthesis or secretion from differentiated lactating mammary cells.

MATERIALS AND METHODS Materials Unless stated otherwise, tissue-culture media were obtained from Gibco Laboratories, Paisley, Renfrewshire, Scotland, U.K., and other reagents were from the Sigma Chemical Co., Poole, Dorset, U.K.

Preparation of mammary epithelial acini from the lactating mouse Mouse mammary epithelial acini were prepared in a similar manner to that described by Turner et al. (1992). Mammary tissue was removed from mid-lactation CD 1 mice (10 days post partum), chopped and dissociated by incubation with collagenase

Abbreviations used: DME, Dulbecco's modified Eagle's; ECM, extracellular matrix; FIL, feedback inhibitor of lactation; HBSS, Hanks balanced salt solution; TGF/3, transforming growth factor-fl. I To whom correspondence should be addressed.

334

Research Communication

(150 i.u./ml; Lorne Laboratories, Reading, U.K.) in Hanks balanced salt solution (HBSS) supplemented with 5 jug/ml insulin, 0.01 jug/ml cortisol, 40 mg/ml BSA, 5.5 mM glucose, 20 #M CaCl2, 10#uM MgSO4, 0.7 mM glutamine and 2 % (v/v) minimum essential amino acids. The suspension was incubated with shaking for up to 90 min at 37 °C until the majority of acini were reduced to clumps of 10-30 cells. Digestion was terminated by filtration of the suspension through a 150 ,um-pore-size nylon mesh (Lockertex, Warrington, Cheshire, U.K.). The acini were harvested by centrifugation at 80 g for 5 min, and resuspended in HBSS supplemented as described above but without BSA or collagenase, and with 14 mM Hepes, 0.04 mg/ml DNAase (Boehringer Mannheim Ltd., Lewes, East Sussex, U.K.) and 0.1 mg/ml soybean trypsin inhibitor. The cells were washed three times and filtered through a 53 ,um-pore-size nylon mesh. The filtrate was centrifuged and the acini were resuspended in culture medium comprising 50 % (v/v) Medium 199, 50 % (v/v) Ham's F12 supplemented with 5,cg/ml insulin, 0.01 ug/ml cortisol, 2.5 mg/ml BSA, 0.1 1ug/ml prolactin, 0.01 jug/ml epidermal growth factor, 0.001 ,ug/ml tri-iodothyronine, 12 mM glucose, 14 mM Hepes and 30 mM NaHCO3. The acini suspension was diluted with culture medium to give an absorbance of 0.015 A450 units/ml, equivalent to 7 x 105 cells/ml.

Incorporation of [NS]methionine Into cellular proteins In a pulse-chase experiment, acini were incubated for 1 h with 20 ,uCi/ml L-[35S]methionine (cell-labelling grade; sp. radioactivity > 1000 Ci/mmol; Amersham p.l.c., Aylesbury, Bucks., U.K.). An equal volume of ice-cold culture medium was added to the acini and the suspension was centrifuged at 80 g for 5 min. The pellet was washed in ice-cold culture medium. In a continuous labelling experiment, acini were incubated with 10 sCi/ml L-[35S]methionine for up to 24 h. Lactating acini were kept in 24-well tissue-culture plates (1 ml/well final volume) at 37 °C under a humidified atmosphere of air/CO2 (19:1). Acini were incubated for various times with recombinant human TGFfl1 (Gibco) or the equivalent volume (100 /1) of vehicle [50 mM sodium acetate (pH 4.5)/1 % (w/v) BSA]. Incorporation of [35S]methionine into cell-associated and secreted trichloroacetic acid-precipitable protein was measured by liquid-scintillation counting as previously described by Turner et al. (1992).

Immunoblottlng of secreted fi-casein Incubations of lactating acini were terminated by centrifugation at 5000 g for 1 min. An equal volume of cold methanol was added to each 500,ul aliquot of supernatant. After 60 min at -70 °C, precipitated protein was pelleted by centrifugation at 5000 g for 2 min. The pellet was washed by centrifugation with cold methanol, solubilized in 100 /tl of SDS dissociation buffer [1.25% (w/v) SDS/2mM EDTA/125mM Tris/10% (w/v) sucrose/l % (v/v) 2-mercaptoethanol/10% (v/v) glycerol/ 0.001 % (w/v) Bromophenol Blue, pH 6.8] and heated to 80 °C for 2 min. Samples were loaded on to 12.5 %-(w/v)-polyacrylamide slab gels and separated by electrophoresis. Purified mouse caseins were run as controls. After electrophoresis, proteins were transferred to nitrocellulose membrane, blots were washed in PBS (138 mM NaCl/3 mM KH2PO4/20 mM Na2HPO4, pH 7.4), incubated for 1 h in PBS with 1 % (w/v) BSA, washed with PBS and incubated for 1 h in polyclonal rabbit anti-(mouse fl-casein) antiserum diluted 1:10000 in PBS. Blots were washed with PBS and incubated for 1 h in peroxidase-labelled anti-rabbit IgG diluted 1: 400 in PBS with 1 % (w/v) BSA and 0.5 % Tween-20. After extensive washing in PBS with 0.5 % Tween-20, immuno-

reactive protein was revealed by enhanced chemiluminescence (ECL; Amersham).

Measurement of [5S]methlonine Incorporated Into proteins secreted from mammary-tissue explants from lactating mice The inguinal mammary glands were removed from a midlactation (10 days post partum) CDI mouse and rinsed in Dulbecco's modified Eagle's (DME)/Ham's F12 mixed medium containing 5 jug/ml gentamycin sulphate. The tissue was minced and 20 explants, weighing 70-80 mg, were placed on to the surface of microfoil plinths in 35 mm-diameter plastic Petri dishes. Each Petri dish was filled with 4 ml of culture medium with 20 ,tCi/ml L-[35S]methionine, and TGF,/1 or vehicle added. Explants were incubated for 3 h at 37 °C under a humidified atmosphere of air/CO2 (19:1). At hourly intervals, 1 ml of medium was removed from each dish and replaced with an equivalent volume of fresh medium (supplemented as described above). Any medium removed was centrifuged at 5000 g for 1 min, and the amount of [35S]methionine incorporated into trichloroacetic acid-precipitable protein measured by liquidscintillation counting as described above.

Immunoblot analysis of proteins secreted from mammary tissue explants from pregnant mice Tissue explants were prepared from the inguinal mammary glands of 15-day-pregnant CD1 mice as described above. Explants were incubated in 4 ml of culture medium comprising DME/Ham's F12 supplemented with 5 ,tg/ml each of gentamycin sulphate, aldosterone, and the lactogenic hormones insulin, cortisol and prolactin (see Robinson et al., 1993). The explants were incubated with TGFf81 or vehicle for 4 days at 37 °C under a humidified atmosphere of air/CO2 (19: 1). Medium was replaced every 24 h and centrifuged at 5000 g for 1 min. Protein present in the medium was precipitated with cold methanol and separated by electrophoresis on 12.5 % gels for immunoblotting as described above. RESULTS In the present study we examined the effect of TGF/?1 on casein synthesis and secretion using three experimental preparations: isolated mammary cell acini from lactating mice, explant cultures of mammary tissue from lactating mice, and mammary-tissue explants from pregnant mice that were allowed to differentiate in culture. Two doses of TGFf81 were chosen. TGFfll at 5 ng/ml has been shown to inhibit maximally casein secretion from mouse mammary explants in a previous study (Robinson et al., 1993), and this concentration, which is equivalent to 200 pM, is within the range of concentrations of TGFf81 (112-236 pM) recently detected in mouse milk (Letterio et al., 1994). TGF/J1 was also used at a supramaximal dose of 50 ng/ml (2 nM). To examine any possible acute effect of TGF/?1 on casein secretion from lactating mammary cells, we made use of a preparation of polarized acini from lactating mouse mammary gland. This preparation has been used extensively in the past to study the characteristics of protein (essentially casein) secretion from mammary epithelial cells (Turner et al., 1992, 1993; Renninson et al., 1993). A pulse-chase labelling protocol allowed the effect of TGFfi1 on secretion of preformed caseins from the lactating acini to be assessed. Previous studies have shown that a 1 h labelling time is sufficient to allow incorporation of label into newly synthesized caseins and some packaging into secretory vesicles, such that secretion begins without a lag during the chase period (Turner et al., 1992). Over a chase period of 3 h, at

335

Research Communication 50e Vehicle 4 -+- 5 ng/ml TGF,B1 ax C.?0 - - 50 ng/mI TGF,t81 a); 40X- 40

25 -

e

u X 20^, o --:0

E O 1C >0 o-X 10.D

l

-o

co

Vehicle 5 ng/ml TGFfl, --x-- 50 ng/mI TGFfl1

0

-2 i 203 00 C

9 x

10 -

cr 0-

O-

0

2

3

Time (h)

Time (h)

Figure 1 Effect of TGFfl1 on secretion of pre-formed protein from lactading mouse mammary acini

Figure 3 Effect of TGFfll on protein secretlon from lactating-mouse mammary Issue explants

Mammary acini were pulse-labelled with 20 ,uCi/ml [35S]methionine for 1 h, then washed and incubated with TGFfl1 (5 or 50 ng/ml) or vehicle. The extent of protein secretion over the following 3 h was determined by measuring radioactivity incorporated into cell-associated or medium trichloroacetic-acid-precipitable protein. Results show the percentages of total radioactivity that were secreted and are expressed as means+ S.E.M. (n = 4).

Mammary-tissue explants from lactating mice were incubated in a continuous-labelling protocol with 20 1c/i/ml [35S]methionine and TGF,f1 (5 or 50 ng/ml) or vehicle. The extent of protein secretion over the following 3 h was determined by measuring the radioactivity incorporated into medium trichloroacetic-acid-precipitable protein. Results are means+ S.E.M. (n = 3).

CD

800- (a)

X

600-

I

e Vehicle + 5 ng/mI TGF,81 --.-- 50 ng/ml TGF,f1

0

.a C._)

o

-s

(3

m

40-

F

20-

x

T10 4-W

i 0O-

-----F o

1

2

3

600- (b)

Q

400-

>

80-

24

I

*° X 60-

X

40-

I-0 0 U

-

0o

Figure 2 Effect of TGFfll acini

2 Time (h) on

3

24

protein synthesis and secretion by lactating-

mouse mammary

Mammary acini were incubated in a continuous-labelling protocol with 10 /ICi/ml [35S]methionine and TGFfl1 (5 or 50 ng/ml) or vehicle. The extent of protein synthesis and secretion over the following 3 h and 24 h later was determined by measuring the radioactive counts incorporated into cell-associated protein (a) or into medium trichloroacetic-acid-precipitable protein (b). Results are means+ S.E.M. (n = 4).

concentrations of 5 or 50 ng/ml, TGFfll did not alter protein secretion from mammary epithelial cells in comparison with cells incubated with vehicle (Figure 1). This result indicated a lack of any direct acute effect of TGF,1 on protein secretion from

lactating acini, but did not preclude the ability of TGF,81 to affect casein synthesis. In a continuous-labelling experiment, isolated mammary acini remained viable in suspension culture and continued to synthesize and secrete protein over a 24 h period. Neither 5 nor 50 ng/ml TGF,81 affected incorporation of label into newly synthesized protein in comparison with cells incubated with vehicle (Figure 2a). Protein secretion from the same cells followed a similar pattern to synthesis (Figure 2b) and was unaffected by TGF,f1 over this time. One explanation for the lack of effect of TGFfl1 on this preparation could be receptor loss during isolation of the acini. Therefore the effect of 5 and 50 ng/ml TGF,f1 on protein secretion from mammary-tissue explants from lactating mice, in which receptors should remain intact, was also examined. TGF/11 did not affect protein secretion over a 3 h period (Figure 3), ruling out this possibility. Immunoblotting of methanol-precipitable protein secreted (over 3 h) from lactating mammary acini, with a polyclonal anti(mouse ,-casein) antiserum detected a single band corresponding to /-casein (Figure 4a). Quantification by scanning densitometry, using purified mouse milk caseins (39 % of which was 8-casein) as the standard, revealed that, over 3 h, lactating mammary acini secreted 774 + 96 ng of 8-casein/106 cells. Caseins represent a small proportion of the total protein secreted from undifferentiated pregnant-mouse mammary tissue; therefore, rather than measuring total protein secreted (as with lactating mammary cells), anti-(ff-casein) antiserum was used to immunoblot protein secreted from mammary-tissue explants from pregnant mice. Results from previous studies suggest that exposure of pregnant-mouse mammary tissue to TGF,81 in vitro over 4 days can inhibit the onset of lactation and subsequent secretion of caseins (Robinson et al., 1993). The effect of the recombinant human TGF,81 preparation used in the present study on f8-casein secretion from pregnant-mouse mammary explants was examined. Mammary explants were incubated with lactogenic hormones and TGF,81 (5 or 50 ng/ml) or vehicle for 4 days. Culture medium was replaced every 24 h and the medium immunoblotted for ,-casein. After 1 or 2 days in culture, immunoblotting of medium from explants detected no, or very small amounts of, ,-casein respectively (results not shown). After 3 or 4 days in culture, fl-casein was detected in medium from explants incubated with vehicle ('- '), but not in medium from explants incubated with 5 or 50 ng/ml TGF,8I

336

Research Communication

who report that

(a) c z

r CN

_ -P:

CY 10

j_i-Casein

(b) Day 3

[TGFft1]

(ng/mi)

-

5

a,,

Day 4 50

-

5

50

u

,)-Casein

Figure 4 Effect of TGF,81 on 8-caseln secretion assessed by immunoblotting (a) An immunoblot which demonstrates that a 1:10000 dilution of polyclonal anti-(mouse ficasein) antiserum detects only 8-casein in methanol-precipitable protein secreted from isolated lactating mammary acini at various times over 3 h. Purified total mouse caseins were included as a control. (b) Immunoblot of methanol-precipitable protein secreted from mammary-tissue explants from pregnant mice. Explants were incubated with lactogenic hormones and TGF/.1 (5 or 50 ng/ml) or vehicle (-) for 4 days. Culture medium was replaced every 24 h and methanol-precipitable protein immunoblotted with a 1:10000 dilution of anti-(mouse /8-casein) antiserum. Results from 3- and 4-day incubations are shown here with total mouse caseins included on the same immunoblot as a control.

(Figure 4b). Thus TGFfll appeared to inhibit the onset of,ficasein production by mammary explants in culture.

DISCUSSION We have demonstrated that, despite the presence of lactogenic hormones, incubation of mammary-tissue explants from midwith 5 5 TGF,#l tor for 3 days 3 or 44 days ng/ml TGFhcu1 pregnant mice with or 50.0 ng/ml prevents the appearance of ,l-casein in the culture medium. This is in agreement with the observations of Robinson et al. (1993), pregnant mice

of explants to

5

ng/ml TGF,f1 (or

or

Received 29 September 1994; accepted 7 October 1994

synthesis and

release at the post-transcriptional level. This latter observation raises the possibility that TGF,/1 may also exert inhibitory effects on production of casein by lactating mammary cells. This hypothesis, however, is not supported by the results presented here. Using a pulse-chase protocol, we have shown that, over a 3 h incubation period, neither 5 nor 50 ng/ml TGF,/1 inhibited secretion of preformed caseins from lactating-mouse mammary epithelial cells. In addition, these concentrations of TGF,/1 were also found not to inhibit casein synthesis or secretion from lactating mammary acini (over 24 h) or casein secretion from lactating mammary-gland explants (over 3 h). Studies have shown that a milk-protein, FIL, suppresses mammary-gland differentiation (Wilde et al., 1990). In addition, FIL acutely inhibits constitutive casein synthesis in, and secretion from, lactating mouse mammary epithelial cells (Renninson et al., 1993), indicating a possible role for FIL in rapid feedback inhibition of casein secretion in vivo. Like FIL, TGF,/1 is present in mouse milk (Letterio et al., 1994) and has been shown to modulate mammary-gland development (Robinson et al., 1991; Jhappan et al., 1993). The results obtained in the present study support the notion that TGF,/l suppresses differentiation of mammary tissue and the subsequent production of 8-casein. The results presented here, however, indicate that an acute role for TGF/31 in inhibition of constitutive synthesis or secretion of caseins from differentiated lactating mammary epithelial cells is unlikely.

~~~~cn

,

exposure

TGF,/2 and TGF,/3) for 4 days inhibits casein

U,

or

This work was funded by a Link Research Grant from the Biotechnology and Biological Sciences Research Council to R.D.B. and C.J.W. and in part by the Scottish Office of the Agricultural Fisheries Department (C.J.W.).

REFERENCES Blatchford, D. R. and Peaker, M. (1982) Q. J. Exp. Physiol. 67, 303-310 Jhappan, C., Geiser, A. G., Kordon, E. C., Bagheri, D., Hennighausen, L., Roberts A. B., Smith, G. H. and Merlino, G. (1993) EMBO J. 12, 1835-1845 Letterio, J. J., Geiser, A. G., Kulkarni, A. B., Roche, N. S., Sporn, M. B. and Roberts, A. B. (1994) Science 264, 1936-1938 Mieth, M., Boehmer, F.-D., Ball, R., Groner, B. and Grosse, R. (1990) Growth Factors 4, 9-15 Renninson, M. E., Kerr, M., Addey, C. V. P., Handel, S. E., Turner, M. D., Wilde, C. J. and Burgoyne, R. D. (1993) J. Cell Sci. 106, 641-648 Robinson, S. D., Silberstein, G. B., Roberts, A. B., Flanders, K. C., and Daniel, C. W. (1991) Development 113, 867-878 Robinson, S. D., Roberts, A. B. and Daniel, C. W. (1993) J. Cell Biol. 120, 245-251 Stampfer, M. R., Yaswen, P., Alhadeft, M. and Hosoda, J. (1993) J. Cell. Physiol. 155,

210-221

Streuli, C. H., Schmidhauser, C., Kobrin, M., Bissel, M. J. and Derynck, R. (1993) J. Cell Biol. 120, 253-260 Turner, D., Renninson, M. E., Handel, S. J. CellM. Biol. 117, 269-278

E. Wilde, C. J. and Burgoyne, R. D. (1992)

Turner, M. D., Handel, S. E., Wilde, C. J. and Burgoyne, R. D. (1993) J. Cell Sci. 106, 1221-1226 Wilde, C. J., Calvert, D. T., Daly, A. and Peaker, M. (1987) Biochem. J. 242, 285-288 Wilde, C. J., Knight, C. H., Addey, C. V. P., Blatchford, D. R., Travers, M., Bennet, C. N. and Peaker, M. (1990) Protoplasma 159, 112-117 Wilde, C. J., Addey, C. V. P., Boddy, L. M. and Peaker, M. (1994) Biochem. J., in the press